Dynamic Resource Allocation for Heterogeneous 5G TDD Wireless Networks: Balancing URLLC and eMBB Services

Efficient resource allocation in 5G heterogeneous networks is critical due to varying quality of service (QoS) requirements. Resource slicing has emerged as a key solution, enabling logical networks with dynamic access to radio resources for improved allocation efficiency. This paper introduces Convex Optimization-Based DTDD (CODTDD), a novel resource allocation scheme for 5G heterogeneous dynamic time division duplex (DTDD) wireless networks. The proposed scheme formulates the resource allocation problem as a through-put maximization problem, constrained by downlink/uplink (DL/UL) isolation, user orthogonality, latency, and fairness requirements. To evaluate the performance of CODTDD, it is benchmarked against two alternative schemes: Search-Assisted Optimization for DTDD (SODTDD) and the conventional round-robin scheduling scheme. SODTDD explores multiple DL/UL durations combined with separate optimizations for DL and UL, while round-robin serves as a baseline comparison. Simulations based on 3GPP Release 16 specifications demonstrate that CODTDD achieves significantly higher throughput than the round-robin scheme while maintaining performance comparable to SODTDD at a fraction of the computational complexity. These results underscore the efficiency and practicality of CODTDD as a resource allocation strategy for 5G DTDD networks.